INDEX

Explanations

technical terms related to digital content, possibly including potential errors

oai_token-act-pair · gpt-3.5-turbo

references to the term "Int" which appears to indicate technical jargon or programming concepts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãģ¦

-0.75

jriwal

-0.71

yards

-0.71

 Gleaming

-0.69

llah

-0.68

theless

-0.67

¶ħ

-0.65

 Confederacy

-0.64

EMENT

-0.64

 compr

-0.63

POSITIVE LOGITS

uitive

1.37

ensity

1.36

ention

1.32

ensive

1.29

ellect

1.29

estinal

1.25

elligence

1.15

ellig

1.04

ENTION

1.01

rast

1.01

Activations Density 0.023%

technical terms related to digital content, possibly including potential errors

references to the term "Int" which appears to indicate technical jargon or programming concepts

No Comments

No Known Activations